LAB #4805 NEUTRAL: LEDGER BENCH: Apply filter: XAUUSD LONG — improves win rate from 31.8% to 53.5% (+21.7pp, n=86) — E8 Intelligence Research

IDEA: Adjustment made to the live model — verify against the backtests: Apply filter: XAUUSD LONG — improves win rate from 31.8% to 53.5% (+21.7pp, n=86) SAME-WINDOW EFFECT: Over the current window, the filter would have blocked the XAUUSD SHORT (rizzy5m, closed LOSS) — but that's a SHORT, and the filter is stated as "XAUUSD LONG," so the mapping is ambiguous. If the filter means "only take XAUUSD longs," it removes one losing trade from the window and leaves the other nine trades untouched; net effect is +1 avoided loss, negligible on a 10-trade sample. If it means "apply this filter to XAUUSD longs specifically," it has zero effect on this window because no XAUUSD long was taken. Either way, the window-level PnL delta is small and the sample is too thin to confirm anything. OVER-TIME PROJECTION: A +21.7pp win-rate jump on n=86 is a large effect from a single-instrument directional filter, which usually signals either a real regime edge or overfitting to a specific XAUUSD trend peri Author: Andrew Stewart Caldin, Independent Researcher, UK. Part of the E8 Intelligence Research series. Platform: e8intelligence.com